The problem is unbounded but only on the second loop

[Asher Metzger]

Hi,

I'm having trouble with a play model I developed.

I get on unbounded exit on the second loop of the Mass_1_linear using linprog. for some reason it seems like it does not even look at the constraints because it is obvious that constraint C5 doesn't work and further more the variables are bounded on other constraints.

When coming in on the second loop the Q_0 variable is no longer only sdpvar but also constants. this is the only difference I can see between the 2 loops.

I would be happy if someone could point me to my error.

Thanks a lot in advance,

Asher

[Johan Löfberg]

>>Mass_1_linear

Undefined function or variable 'mass'.

Error in Mass_1_linear (line 13)

    [Q_0,cost] = mass(aa,years,ii);

 

>> 

[Asher Metzger]

sorry. wrong file.

now it should run.

[Johan Löfberg]

It is unbounded in all iterations except the first (using a real solver, such as gurobi, would never trust linprog to judge this)

You can see the huge solution, by artificially constraining it to something large but finite

sol2   = optimize([Const2,obj(ii) >= -10000],obj(ii),ops1);
value(Mp)value(Md)
...